Real Time Gesturing in Embodied Agents for Dynamic Content Creation

Hazel Watson-Smith (Soul Machines), Felix Marcon Swadel (Soul Machines), Jo Hutton (Soul Machines), Kirstin Marcon (Soul Machines), Mark Sagar (Soul Machines), Shane Blackett (Soul Machines), Tiago Rebeiro (Soul Machines), Travers Biddle (Soul Machines), Tim Wu (Soul Machines)

Abstract

The content creation industry is experiencing significant growth and the utilisation of Real-Time Gesturing in embodied agents presents an excellent opportunity to enhance the communication of text. Using the proposed system, raw text can be parsed in realtime and an appropriate emotional and gestural performance is generated. It can also be configured to convey personality traits using elements such as emotional state and responses to stimuli, gesture rate, type, size and speed, and augmented with inserted markup tags.